A spline based regression technique on interval valued noisy data

In this paper we present a spline based center and range method (SCRM) to perform regression on interval valued noisy data. The method provides a fast and accurate mechanism to model and predict upper and lower limits of unknown functions in a bounded design space. This technique is superior to previously existing techniques like center and range linear least square regression (CRM). The accurate models may find wide usage in high precision applications. The effectiveness of the proposed technique is demonstrated through experiments on datasets with various applications.